Big Success: Pioneering a free, ad-supported telephone service that provides callers with movie times in 30 cities and handles 2.5 million inquiries a week.
Smart Mistake: Launching the company's advance ticketing service with a low-tech fulfillment system.
Back in 1991, when we started to sell advance tickets to movies, we set up the lowest-tech system imaginable. It was basically a PC, a modem, and a printer in the box office of the Pacific Cinerama Dome, an 1,100-seat theater in Los Angeles. We'd ask callers for their credit-card number, the movie they wanted to see, and the number of tickets. Then we'd send a message to the printer at the theater. The printer was loaded with "will call" envelopes. Someone would see a message -- "two tickets for Aladdin at 7 PM" -- and then print them out and stick them in the envelope.
It was a disaster. We launched the service right when 'Terminator 2" opened. On the first day, theater employees came in and had to wade into the box office. The floor was covered with envelopes. They had to print out the advance tickets at the same time they sold tickets to people who walked up to the window.
It was the scrappy way to do it. But it was a nightmare. We should have done it in a methodical way -- tested the system, done some forecasting. But our "mistake" also illustrated quite graphically that there was a demand for this service, and that we were the ones to handle it. Today we have a sophisticated system, with a machine that sits on the wall at theaters, debits tickets directly out of inventory, and prints them out in an automated way. That's the high-tech solution.
Big Success: Inventing the market for online city guides; expanding from 15 employees in 1995 to 500 today.
Smart Mistake: Committing to the wrong technology.
In 1995, when we started CitySearch, Java had just been invented. It was going to revolutionize the Internet. So we created the first full-scale commercial Java application anywhere. It was painstaking to build. It occupied 15 programmers for four months. And it was huge, requiring 18 megabytes to download.
We launched the application in May 1996. We also launched a more traditional HTML version, almost as an afterthought. Well, it turns out that fewer than 10% of our users ever bothered to use the stand-alone Java application; it took too long to download. We no longer even support it.
The project was a great disappointment. But it was a success in other ways. It gave us lots of expertise in Java, and that has given us a huge leg-up as Java applets have become more important. It also helped with raising money. Investors wanted to see whether we could do cool technology, and Java was definitely cool.
It also taught me an important lesson: long lead times compound mistakes. If you keep your development time short and go to prototype quickly, you can make mistakes and correct them before they sink the company. The future never unfolds the way you think it will.
Big Success: Establishing the dominant brand in online bookselling; taking the company public in May 1997.
Smart Mistake: Overcommitting on startup inventory.
My smartest mistake wound up playing a big role in our success. When we launched Amazon.com in July 1995, we announced that we would offer a million book titles. A rational, deliberate decision would have been to offer maybe 500,000 titles and then ramp up as we learned how to fulfill orders. But operating from a marketing and branding point of view, we said, "Damn the torpedoes."
We developed a fulfillment plan, but it was a huge gamble. We weren't sure it would work. If we had messed up, our customers would have spread the word over the Net to hundreds of thousands of people. It worked out, but it took a huge amount of focus and energy.
I really believe the decision was a mistake. We got lucky. I would still advise people to make the rational, deliberate choice the vast majority of the time.
Email
Digg
StumbleUpon
Facebook
Buzz Up!
LinkedIn
